Lung Ultrasound in Bronchopulmonary Dysplasia: Diagnostic Tool, Prognostic Marker or Monitoring Strategy?

Bronchopulmonary dysplasia (BPD) remains one of the leading chronic respiratory complications of extreme prematurity despite major advances in neonatal intensive care. Current diagnostic definitions are primarily based on respiratory support requirements and provide limited information on the biological heterogeneity and longitudinal evolution of lung injury. Lung ultrasound (LUS) is progressively reshaping the clinical approach to BPD by enabling radiation-free, bedside, and repeatable assessment of peripheral lung abnormalities throughout the neonatal course. This narrative review examines the current role of LUS in BPD, integrating its pathophysiological basis with the available evidence and distinguishing explicitly between its diagnostic, prognostic, monitoring, and treatment-guiding applications, which are supported by markedly different levels of evidence. We discuss how LUS findings reflect the major components of BPD pathophysiology, compare LUS with conventional imaging modalities, and summarize the scanning protocols, semiquantitative scoring systems, examination schedules, and patient populations that have been evaluated to date. We also review the emerging contribution of artificial intelligence to automated image analysis, standardized image acquisition, and personalized risk prediction. Most of the available evidence is prognostic rather than diagnostic: early LUS scores predict BPD subsequently defined at 36 weeks' postmenstrual age, whereas LUS has not been validated as a diagnostic test for established BPD, and reported cutoffs remain study- and protocol-specific rather than universally applicable. Randomized evidence is confined to LUS-guided surfactant administration, where the demonstrated benefits concern the timing of treatment and the need for invasive ventilation; no trial has yet shown that LUS-guided management reduces the incidence of BPD. Current evidence supports LUS as a complementary imaging modality that extends beyond the assessment of acute neonatal respiratory disease and enables longitudinal bedside monitoring of peripheral lung aeration. Although further multicenter prospective studies are required to harmonize protocols and validate LUS-guided management strategies, the integration of standardized LUS assessment with emerging artificial intelligence technologies has the potential to establish LUS as a key component of precision respiratory care for infants at risk of BPD.